War-game a customer health score before inputs, thresholds, and playbooks harden into how every renewal and escalation gets triaged.
Customer health scores fail when proxies invent risk the account never showed, when red accounts still have no owner, when sales overrides the score without a log, and when customers hear a number they were never meant to see. A neat dashboard tile is not evidence.
Freeze the package
State which signals feed the score, how weights are set, who owns red and yellow playbooks, what success looks like after two quarters of renewals, and the kill criteria if the score never changes an action. Attach the signal dictionary, sample accounts, playbook owners, and the privacy constraints on sharing scores externally. If CS, sales, and product disagree on whether usage dips mean risk or seasonal quiet, reconcile before seating.
Name the decision you will make if the war game finds nothing new, and the hold criteria if any red threshold lacks a named owner or a measured playbook close rate.
Seats
- Customer success lead. Playbooks that look complete while capacity cannot clear the red queue.
- Account executive. Fear that a red score becomes blame in a forecast call.
- Product. Signals that punish missing features customers already filed.
- RevOps or analytics. Weighting fiction and dual-counted usage metrics.
- Skeptic. The claim that looks strongest and is least sourced.
Optional counsel seat if scores will be shared with customers or used in contractual remedies. Same pack for every seat.
Run the loop
Feed Pingpong the score memo and exhibits. First pass steelmans the model. Later passes attack from the seats above. Final pass turns surviving objections into fairer signals, clearer playbook owners, or a hold. Delete invented "we already know who will churn" claims and dual-counted outreach hours.
Force a day-after narrative: what happens if a strategic account goes red overnight on a noisy proxy, if sales overrides without a log, or if a customer screenshots an internal score. If those stories are stronger than your playbook plan, fix the package before you operationalize it. Separate a health score from a churn playbook or a renewal playbook so each decision stays testable.
Ask every seat to mark which steps are optional in practice. Optional weekly reviews of red accounts become skipped within a month if nobody names the owner and the time box. Related: war-game a customer success playbook, war-game a churn playbook, war-game a renewal playbook, pretend you are the customer success lead, and the war-game decisions hub. Process: how to run a Pingpong.